The current surgical peer review process, most commonly in the form of the Mortality and Morbidity (M&M) conferences, focuses on assessing clinical complications for opportunities of improvement. However, such a framework may overlook lessons and insights that may be available from cases without complications, especially cases with positive outcomes despite being high risk for postoperative complications. These cases with unexpected positive outcomes have been referred to as positive deviants. The concept of positive deviance is being increasingly applied to quality improvement within healthcare organizations, but has rarely been used in the context of M&M.1
A major barrier to implementing positive deviance to M&M case reviews is how to identify unexpected clinical success cases and how to allocate discussion of such cases during a conference that may already be constrained for time. Given the relatively low complication rates in modern medicine, with substantially more patients experiencing positive than negative outcomes, a system that attempts to review all positive cases will be easily overwhelmed.2 Moreover, not all positive outcomes provide valuable lessons, as many patients with good outcomes are low risk and are not expected to have poor outcomes. Only cases with unexpected outcomes, or those who did well despite a high-risk profile, may provide valuable insights.3 Therefore, an objective approach to determine patient risk and to differentiate positive outcomes between those that were expected versus those that were unexpected is necessary to implement the new framework.
Risk calculators, in conjunction with the advent of large population databases, have been developed to assess the likelihood of negative outcomes for any given patient given their unique combination of characteristics.4,5 One such risk calculator is based on the American College of Surgeons National Surgical Quality Improvement Program (NSQIP),6 which has been demonstrated to be useful in aiding discussion between surgeons and patients.7 We hypothesize that this risk assessment tool can objectively identify positive deviant cases prior to M&M conferences. As such, we aim to demonstrate the application of NSQIP risk calculators to devise a system to objectively identify positive deviant cases for qualitative discussion in M&M conferences.
A retrospective chart review of adult complex spine surgery cases from 2013 to 2017 at a single hospital was performed. Patients were included if they underwent posterior spine instrumentation or spinal decompression identified via Current Procedure Terminology codes 22840, 22842, 22843, 22844, and 63005. The risks of postoperative complications were computed utilizing the NSQIP online calculator,6 based on each patient’s demographics, comorbidities, and primary procedure.
A case with an unexpected success, or positive deviant, was defined as having 1) a calculated risk of serious complication greater than that of the population average, but without experiencing a serious complication and 2) a calculated risk of serious complication above a natural break in the distribution of patient data during the study period (Fig. 1). Serious complications examined in this study included cardiac complication, acute renal failure, renal insufficiency, pneumonia, pulmonary embolism, deep vein thrombosis, systemic sepsis, unplanned intubation, unplanned return to the operating room, urinary tract infection, surgical site infection, and wound disruption.6
The study population included 379 patients with an average age of 64.3 years and the majority of whom were American Society of Anesthesiologists Physical Status Class I–II (Table 1). There were 339 patients without a postoperative serious complication (90.6%). The calculated risk for each case without an observed serious complication ranged from 0.4% to 18.4%; while the population average of serious complication reported by the NSQIP risk calculator was 5.3% for procedures that involved less than 3 spine levels, 8.4% for procedures that involved 3 to 6 spine levels, 14.9% for procedures that involved 7 to 12 spine levels, 13.1% for procedures that involved 13 or more spine levels, and 6.1% for spinal decompression procedures.
TABLE 1. -
Patient Demographics and Clinical Characteristics
|Number of patients
||64.3 ± 12.2
|Hypertension requiring medication
|ASA physical status class
| PSI for less than 3 spine levels
| PSI for 3–6 spine levels
| PSI for 7–12 spine levels
| PSI for 13 spine levels or more
| Spinal decompression
|Patients without an observed serious complication
|Patients had an above-average calculated risk but without an observed serious complication
|Calculated risk for patients without an observed serious complication
||7.2% ± 2.6% (range: 0.4%–18.4%)
Data are illustrated in numbers (percentage) or mean ± standard deviation.
ASA indicates American Society of Anesthesiologists; PSI, posterior spine instrumentation.
Among the study population, 13.7% (n = 52) did not experience a serious complication despite having an above-average expected risk. Based on the proposed framework, one of these cases could be considered an unexpected success, or a positive deviant (Fig. 1). Notably, this patient had a calculated risk of serious complication of 18.4%, but did not experience a serious complication.
To enhance M&M discussion, we describe the application of a positive deviance framework and the process of identifying unexpected clinical success. Leveraging the NSQIP risk calculator, we introduce an objective method to identify cases, which improves upon the traditional approach of subjective nomination based on personal judgment. Moreover, the NSQIP risk data facilitates algorithmic prioritization by rank ordering cases for discussion, thus optimizing time management and the educational return per unit time. For the discussion of unexpected success, we rank cases without complications from higher risk to lower risk to help prioritize the unexpected cases. Similarly, this approach can be used to identify unexpected complications; and in those cases, we would rank cases with complications from lower risk (more unexpected in these cases) to higher risk.
In addition to algorithmic identification and prioritization of cases, the NSQIP risk calculator can further provide contextual national data to augment the actual case discussions. While literature reviews may supplement discussion, their data may be limited in generalizability; in contrast, the NSQIP risk calculator yields an objective estimate of the expected risk for each patient and compares that risk to the average risk among similar patients nationwide. A discussion about a high-risk patient that did not experience a complication might focus on rehabilitation, preoperative risk optimization, intraoperative decision making, postoperative care, early mobilization, well-planned discharge, and appropriate outpatient follow-up.
In conclusion, a review of positive deviants may add valuable insight to improve the peer-review process. The proposed framework has important implications as it lays the foundation for subsequent qualitative reviews of positive deviants to help better understand and learn from those cases. The culture of M&M has evolved from identifying individual errors for the goal of quality improvement to becoming an educational venue for faculty and trainees.7,8 If the aim of M&M is to provide systems-based learning opportunities, then adding the selective review of positive outliers may enhance educational value compared to the traditional focus of discussing only complications. Furthermore, the framework can help boost the morale of the healthcare team by highlighting positive contributions of surgeons, collaborating clinicians, and the hospital system.
1. Baxter R, Taylor N, Kellar I, et al. What methods are used to apply positive deviance within healthcare organisations? A systematic review. BMJ Qual Saf. 2016;25:190–201.
2. Chang DC, Yu PT, Easterlin MC, et al. Demystifying sample-size calculation for clinical trials and comparative effectiveness research: the impact of low-event frequency in surgical clinical research. Surg Endosc. 2013;27:359–363.
3. Bohnen JD, Chang DC, Lillemoe KD. Reconceiving the morbidity and mortality conference in an era of big data: an “unexpected” outcomes approach. Ann Surg. 2016;263:857–859.
4. Spence RT, Chang DC, Chu K, et al. An online tool for global benchmarking of risk-adjusted surgical outcomes. World J Surg. 2017;41:24–30.
5. Spence RT, Mueller JL, Chang DC. A novel approach to global benchmarking of risk-adjusted surgical outcomes: beyond perioperative mortality rate. JAMA Surg. 2016;151:501–502.
6. ACS Risk Calculator. Available at: https://riskcalculator.facs.org/RiskCalculator/about.html
. Accessed October 14, 2021.
7. Hutter MM, Rowell KS, Devaney LA, et al. Identification of surgical complications and deaths: an assessment of the traditional surgical morbidity and mortality conference compared with the American College of Surgeons-National Surgical Quality Improvement Program. J Am Coll Surg. 2006;203:618–624.
8. Anderson JE, Jurkovich GJ, Galante JM, et al. A survey of the surgical morbidity and mortality conference in the United States and Canada: a dying tradition or the key to modern quality improvement? J Surg Educ. 2021;78:927–933.